FieldScreen, a ligand-based Virtual Screening (VS) method, is described. Its use of 3D molecular fields makes it particularly suitable for scaffold hopping, and we have rigorously validated it for this purpose using a clustered version of the Directory of Useful Decoys (DUD). Using thirteen pharmaceutically relevant targets, we demonstrate that FieldScreen produces superior early chemotype enrichments, compared to DOCK. Additionally, hits retrieved by FieldScreen are consistently lower in molecular weight than those retrieved by docking. Where no X-ray protein structures are available, FieldScreen searches are more robust than docking into homology models or apo structures.

In this review, we highlight recent applications of machine learning to virtual screening, focusing on the use of supervised techniques to train statistical learning algorithms to prioritize databases of molecules as active against a particular protein target. Both ligand-based similarity searching and structure-based docking have benefited from machine learning algorithms, including naïve Bayesian classifiers, support vector machines, neural networks, and decision trees, as well as more traditional regression techniques. Effective application of these methodologies requires an appreciation of data preparation, validation, optimization, and search methodologies, and we also survey developments in these areas.

We present a comparative assessment of several state-of-the-art machine learning tools for mining drug data, including support vector machines (SVMs) and the ensemble decision tree methods boosting, bagging, and random forest, using eight data sets and two sets of descriptors. We demonstrate, by rigorous multiple comparison statistical tests, that these techniques can provide consistent improvements in predictive performance over single decision trees. However, within these methods, there is no clearly best-performing algorithm. This motivates a more in-depth investigation into the properties of random forests. We identify a set of parameters for the random forest that provide optimal performance across all the studied data sets. Additionally, the tree ensemble structure of the forest may provide an interpretable model, a considerable advantage over SVMs. We test this possibility and compare it with standard decision tree models.

Chemotype enrichment is increasingly recognized as an important measure of virtual screening performance. However, little attention has been paid to producing metrics which can quantify chemotype retrieval. Here, we examine two different protocols for analyzing chemotype retrieval: "cluster averaging", where the contribution of each active to the scoring metric is proportional to the number of other actives with the same chemotype, and "first found", where only the first active for a given chemotype contributes to the score. We demonstrate that this latter analysis, common in the qualitative analysis used in the current literature, has important drawbacks when combined with quantitative metrics.